2024-04-15 18:01:55 +00:00
|
|
|
FROM webrecorder/browsertrix-crawler:1.0.4
|
2022-10-31 17:10:55 +00:00
|
|
|
|
2023-05-09 10:12:02 +00:00
|
|
|
ENV RUNNING_IN_DOCKER=1
|
|
|
|
|
|
2023-09-08 14:09:50 +00:00
|
|
|
WORKDIR /app
|
2022-10-31 17:10:55 +00:00
|
|
|
|
|
|
|
|
RUN pip install --upgrade pip && \
|
|
|
|
|
pip install pipenv && \
|
2023-05-10 12:29:42 +00:00
|
|
|
add-apt-repository ppa:mozillateam/ppa && \
|
2022-10-31 17:10:55 +00:00
|
|
|
apt-get update && \
|
2023-07-28 11:46:30 +00:00
|
|
|
apt-get install -y gcc ffmpeg fonts-noto exiftool && \
|
2023-05-10 12:29:42 +00:00
|
|
|
apt-get install -y --no-install-recommends firefox-esr && \
|
|
|
|
|
ln -s /usr/bin/firefox-esr /usr/bin/firefox && \
|
|
|
|
|
wget https://github.com/mozilla/geckodriver/releases/download/v0.33.0/geckodriver-v0.33.0-linux64.tar.gz && \
|
2022-11-02 16:51:32 +00:00
|
|
|
tar -xvzf geckodriver* -C /usr/local/bin && \
|
|
|
|
|
chmod +x /usr/local/bin/geckodriver && \
|
2023-05-10 12:29:42 +00:00
|
|
|
rm geckodriver-v*
|
2022-11-08 13:59:35 +00:00
|
|
|
|
|
|
|
|
|
2023-05-09 15:38:17 +00:00
|
|
|
COPY Pipfile* ./
|
2023-08-24 15:57:58 +00:00
|
|
|
# install from pipenv, with browsertrix-only requirements
|
2024-04-15 18:01:55 +00:00
|
|
|
RUN pipenv install
|
|
|
|
|
|
2023-05-09 16:45:02 +00:00
|
|
|
# doing this at the end helps during development, builds are quick
|
2022-11-08 15:55:33 +00:00
|
|
|
COPY ./src/ .
|
2022-10-31 17:10:55 +00:00
|
|
|
|
2023-05-10 09:08:49 +00:00
|
|
|
ENTRYPOINT ["pipenv", "run", "python3", "-m", "auto_archiver"]
|
2023-05-09 15:38:17 +00:00
|
|
|
|
2023-05-10 08:51:53 +00:00
|
|
|
# should be executed with 2 volumes (3 if local_storage is used)
|
2023-09-14 16:49:37 +00:00
|
|
|
# docker run --rm -v $PWD/secrets:/app/secrets -v $PWD/local_archive:/app/local_archive aa pipenv run python3 -m auto_archiver --config secrets/orchestration.yaml
|
